If accuracy is already strong, analyze where the time is going. Slow performance often comes from overchecking, rereading, or spending too long comparing close options rather than calculation speed.

Set time targets for the parts that repeatedly expand. Practice committing once the evidence is strong enough instead of trying to remove every uncertainty.

Protect accuracy by changing the process, not by simply forcing yourself to move faster. If useful, this guide on McKinsey Solve guide gives more detail.

sounds like overchecking more than speed problem.

I set a max time for close-call decisions and force myself to commit.